Can machine learning on learner analytics produce a predictive model on student performance?

The aim of this research is to analysis past student learner analytics using machine learning algorithms that had undertaken a web development and programming module. By specifically using the access and error web server logs from each student web server it provides a deeper learner analytic data. The web server logs every web file access and error access from a browser so in turn each data file can directly relate to a student's engagement level and assessment strategy. Each log holds several types of information which is filtered to make sure only the relevant dataset is processed through the machine learning framework. The students' performance data was also gathered so that a data mining analysis the learner analytics could be performed to see if there is any correlation between log data and their final assessment mark.

WEKA, an open source machine learning software suite was used to perform data mining algorithms on the large data set. Applying data mining in education is an emerging research field also known as educational data mining (EDM) and some studies have found that EDM could predict with a success rate of more than 80% which students will or will not graduate. By using data mining on student's assignment development data it would show a correlation and therefore a predictive model could be produced.